CN116843345A - Intelligent wind control system and method for trading clients based on artificial intelligence technology - Google Patents
Intelligent wind control system and method for trading clients based on artificial intelligence technology Download PDFInfo
- Publication number
- CN116843345A CN116843345A CN202311101297.6A CN202311101297A CN116843345A CN 116843345 A CN116843345 A CN 116843345A CN 202311101297 A CN202311101297 A CN 202311101297A CN 116843345 A CN116843345 A CN 116843345A
- Authority
- CN
- China
- Prior art keywords
- transaction data
- training
- historical
- semantic
- feature vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000005516 engineering process Methods 0.000 title claims abstract description 41
- 239000013598 vector Substances 0.000 claims abstract description 146
- 230000002159 abnormal effect Effects 0.000 claims abstract description 30
- 238000012549 training Methods 0.000 claims description 88
- 238000012546 transfer Methods 0.000 claims description 38
- 239000011159 matrix material Substances 0.000 claims description 35
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 9
- 230000007246 mechanism Effects 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 3
- 230000006399 behavior Effects 0.000 description 19
- 238000010586 diagram Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 9
- 230000008569 process Effects 0.000 description 8
- 238000009826 distribution Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012502 risk assessment Methods 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/38—Payment protocols; Details thereof
- G06Q20/40—Authorisation, e.g. identification of payer or payee, verification of customer or shop credentials; Review and approval of payers, e.g. check credit lines or negative lists
- G06Q20/401—Transaction verification
- G06Q20/4016—Transaction verification involving fraud or risk level assessment in transaction processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Computer Security & Cryptography (AREA)
- Economics (AREA)
- Marketing (AREA)
- Technology Law (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
An intelligent wind control system and method for client transaction based on artificial intelligence technology is disclosed. Firstly, acquiring historical transaction data of a served object, then, acquiring current transaction data of the served object, then, carrying out semantic coding on the historical transaction data to obtain a historical transaction data semantic understanding feature vector, then, carrying out semantic coding on the current transaction data to obtain a current transaction data semantic understanding feature vector, and finally, generating a wind control result based on the current transaction data semantic understanding feature vector and the historical transaction data semantic understanding feature vector. In this way, the historical transaction behavior of the client can be learned based on the artificial intelligence technology, and the recent abnormal transaction behavior can be timely identified and followed.
Description
Technical Field
The application relates to the field of intelligent wind control, in particular to an intelligent wind control system and method for trading clients based on an artificial intelligence technology.
Background
In the field of financial investments, it is critical for financial institutions and investors to know the historical transaction behavior of customers and to identify and follow up in time recent abnormal transaction behavior. By analyzing the historical transaction data of the customer, the financial institution can be assisted in identifying potential risks and taking corresponding measures to protect the interests of the customer and maintain the stability of the market.
Thus, a wind control scheme for customer transactions is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides an intelligent wind control system and method for trading clients based on an artificial intelligence technology. The method can learn the historical transaction behaviors of the clients based on artificial intelligence technology, and timely identify and follow up recent abnormal transaction behaviors.
According to one aspect of the present application, there is provided an intelligent wind control system for trading customers based on artificial intelligence technology, comprising:
the historical transaction data acquisition module is used for acquiring historical transaction data of the served object, wherein each transaction data in the historical transaction data comprises transaction time, transaction variety, transaction direction, transaction price, transaction quantity, transaction cost, transaction type, transaction result, transaction state and transactor information;
the current transaction data acquisition module is used for acquiring current transaction data of the served object;
the historical data semantic coding module is used for carrying out semantic coding on the historical transaction data to obtain a historical transaction data semantic understanding feature vector;
the current data semantic coding module is used for carrying out semantic coding on the current transaction data to obtain a semantic understanding feature vector of the current transaction data; and
and the wind control result generation module is used for generating a wind control result based on the semantic understanding feature vector of the current transaction data and the semantic understanding feature vector of the historical transaction data.
According to another aspect of the present application, there is provided an intelligent wind control method for a customer transaction based on artificial intelligence technology, comprising:
acquiring historical transaction data of a served object, wherein each transaction data in the historical transaction data comprises transaction time, transaction variety, transaction direction, transaction price, transaction quantity, transaction cost, transaction type, transaction result, transaction state and transactor information;
acquiring current transaction data of a served object;
carrying out semantic coding on the historical transaction data to obtain a semantic understanding feature vector of the historical transaction data;
carrying out semantic coding on the current transaction data to obtain semantic understanding feature vectors of the current transaction data; and
and generating a wind control result based on the current transaction data semantic understanding feature vector and the historical transaction data semantic understanding feature vector.
Compared with the prior art, the intelligent wind control system and the method for trading clients based on the artificial intelligence technology are characterized in that firstly, historical trading data of a served object is obtained, then, current trading data of the served object is obtained, then, semantic encoding is carried out on the historical trading data to obtain a semantic understanding feature vector of the historical trading data, then, semantic encoding is carried out on the current trading data to obtain the semantic understanding feature vector of the current trading data, and finally, wind control results are generated based on the semantic understanding feature vector of the current trading data and the semantic understanding feature vector of the historical trading data. In this way, the historical transaction behavior of the client can be learned based on the artificial intelligence technology, and the recent abnormal transaction behavior can be timely identified and followed.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly introduced below, the following drawings not being drawn to scale with respect to actual dimensions, emphasis instead being placed upon illustrating the gist of the present application.
FIG. 1 is a block diagram of an intelligent wind control system for trading customers based on artificial intelligence technology in accordance with an embodiment of the present application.
FIG. 2 is a block diagram of the historical data semantic encoding module in an intelligent wind control system based on artificial intelligence technology for trading customers according to an embodiment of the present application.
FIG. 3 is a block diagram of the wind control result generation module in the intelligent wind control system for trading customers based on artificial intelligence technology according to the embodiment of the application.
FIG. 4 is a block diagram schematic of a training module further included in an intelligent wind control system for trading customers based on artificial intelligence techniques in accordance with an embodiment of the present application.
FIG. 5 is a flow chart of an intelligent wind control method for a customer transaction based on artificial intelligence technology in accordance with an embodiment of the present application.
FIG. 6 is a schematic diagram of a system architecture of an intelligent wind control method for customer transactions based on artificial intelligence technology according to an embodiment of the present application.
FIG. 7 is an application scenario diagram of an intelligent wind control system for customer transactions based on artificial intelligence techniques according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are also within the scope of the application.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Although the present application makes various references to certain modules in a system according to embodiments of the present application, any number of different modules may be used and run on a user terminal and/or server. The modules are merely illustrative, and different aspects of the systems and methods may use different modules.
A flowchart is used in the present application to describe the operations performed by a system according to embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
Aiming at the technical requirements, the technical conception of the application is as follows: based on artificial intelligence technology, learning historical transaction behaviors of clients, and timely identifying and following recent abnormal transaction behaviors. It should be appreciated that from the historical transaction behavior data may reflect the user's transaction style and transaction characteristics, and that, based on comparing recent transaction data with the customer's historical transaction patterns, abnormal transaction behavior may be identified, including: a. abnormal transaction amount: the customers conduct large-amount transactions in a short time, which exceeds the normal transaction amount range; b. abnormal transaction mode: the transaction behavior of the customer changes suddenly and is not consistent with the historical transaction mode. For example, clients move from robust investment strategies to high-risk investments; abnormal transaction time: the customer makes transactions during non-transaction periods or frequently for short periods of time.
Further, once abnormal transaction behavior is identified, the financial institution needs to take timely measures to follow up, comprising the following steps: a. verifying abnormal behaviors: first, the financial institution needs to verify whether there is indeed an abnormal transaction behaviour. This can be achieved by further data analysis and communication with the customer; b. contacting the client: if abnormal transaction behaviors are confirmed, the financial institution should contact the clients, know the transaction intention of the clients and provide necessary explanation and advice; c. risk management measures: depending on the nature and the degree of risk of the abnormal transaction, the financial institution may take appropriate risk management measures, such as limiting the transaction amount, suspending the transaction account, etc.; monitoring and evaluating: the financial institution should continue to monitor the customer's transaction and evaluate whether the action taken is valid. If abnormal transaction activity persists or reappears, further investigation and more stringent action may be required.
FIG. 1 is a block diagram of an intelligent wind control system for trading customers based on artificial intelligence technology in accordance with an embodiment of the present application. As shown in fig. 1, an intelligent wind control system 100 for trading customers based on artificial intelligence technology according to an embodiment of the present application includes: a historical transaction data obtaining module 110, configured to obtain historical transaction data of a served object, where each transaction data in the historical transaction data includes a transaction time, a transaction variety, a transaction direction, a transaction price, a transaction quantity, a transaction fee, a transaction type, a transaction result, a transaction state and transaction information; a current transaction data acquisition module 120, configured to acquire current transaction data of a served object; the historical data semantic coding module 130 is configured to perform semantic coding on the historical transaction data to obtain a historical transaction data semantic understanding feature vector; the current data semantic coding module 140 is configured to perform semantic coding on the current transaction data to obtain a semantic understanding feature vector of the current transaction data; and a wind control result generating module 150, configured to generate a wind control result based on the current transaction data semantic understanding feature vector and the historical transaction data semantic understanding feature vector.
Specifically, in the technical scheme of the application, firstly, historical transaction data of a served object and current transaction data of the served object are obtained, wherein each transaction data in the historical transaction data and the current transaction data comprise transaction time, transaction variety, transaction direction, transaction price, transaction quantity, transaction cost, transaction type, transaction result, transaction state and transactor information.
And then, carrying out semantic coding on the historical transaction data to obtain a semantic understanding feature vector of the historical transaction data, and carrying out semantic coding on the current transaction data to obtain a semantic understanding feature vector of the current transaction data. And generating a wind control result after obtaining the semantic understanding feature vector based on the current transaction data and the semantic understanding feature vector based on the historical transaction data.
More specifically, in one specific example of the present application, the process of semantically encoding the historical transaction data to obtain the semantic understanding feature vector of the historical transaction data includes: firstly, each transaction data in the historical transaction data is respectively passed through a semantic encoder comprising an embedded layer to obtain a plurality of transaction data semantic encoding feature vectors. That is, in the process of semantically encoding historical transaction data, semantic encoding is performed on each transaction data in the historical data to obtain a plurality of transaction data semantic encoding feature vectors. Further, the plurality of transaction data semantically encoded feature vectors are passed through a historical transaction data feature extractor based on a converter module to obtain the historical transaction data semantically understood feature vector. That is, after obtaining the semantic coding feature representation of each transaction data in the historical transaction data, performing context-dependent coding based on a vector sequence on the semantic coding feature representation of each transaction data to obtain the semantic understanding feature vector of the historical transaction data.
Accordingly, as shown in fig. 2, the historical data semantic encoding module 130 includes: a historical data embedding encoding unit 131, configured to obtain a plurality of transaction data semantic encoding feature vectors by respectively passing each transaction data in the historical transaction data through a semantic encoder including an embedding layer; and a historical data feature extraction unit 132 for passing the plurality of transaction data semantically encoded feature vectors through a historical transaction data feature extractor based on the converter module to obtain the historical transaction data semantically understood feature vector. It should be appreciated that the embedding layer is a hierarchical structure in a deep learning model for converting discrete data (e.g., text, images, etc.) into a continuous vector representation that can map high-dimensional sparse input data to a low-dimensional dense feature space, thereby capturing semantic relationships and similarities between the data. Each transaction data in the history transaction data is encoded by the above-described history data embedding encoding unit 131 using an embedding layer, and is converted into a transaction data semantic encoding feature vector. The purpose of this is to represent the transaction data in a continuous vector form for subsequent feature extraction and processing. By using an embedded layer, the original historical transaction data can be converted into a vector representation with semantic information so that the model can better understand and process the data. The semantic coding feature vector can be used for subsequent tasks such as data analysis, pattern recognition, prediction and the like, and the performance and effect of the model are improved. The converter module is a module for feature conversion and extraction, commonly used in deep learning models, and functions to convert input data into a more useful and meaningful representation through a series of transformations and operations. The historical transaction data feature extractor based on the converter module is used by the historical data feature extraction unit 132 to process the plurality of transaction data semantic coding feature vectors to obtain the historical transaction data semantic understanding feature vectors. The converter module is typically composed of multiple layers, such as convolutional layers, recurrent neural network layers, attention mechanisms, and the like. These layers may perform various operations on the input data, such as filtering, pooling, sequence modeling, etc., to extract key features and patterns in the data. By using the converter module, the original transaction data semantically encoded feature vector can be processed and converted more deeply, so that a higher-level feature representation is extracted. These feature representations can better capture semantic information and important features of the data, providing more useful input for subsequent tasks (e.g., classification, prediction, etc.).
More specifically, in one specific example of the present application, the process of semantically encoding the current transaction data to obtain the semantically understood feature vector of the current transaction data includes: the current transaction data is passed through a converter module-based current transaction data feature extractor comprising an embedded layer to obtain the current transaction data semantic understanding feature vector. That is, the current transaction data is semantically encoded using the converter module-based current transaction data feature extractor comprising an embedded layer to obtain the current transaction data semantically understood feature vector. The method comprises the steps of firstly converting each data item in the current transaction data into a plurality of current transaction data item embedded vectors through the embedding layer, and then carrying out context semantic coding based on a multi-head attention mechanism on the plurality of current transaction data item embedded vectors by using a current transaction data feature extractor based on a converter module to obtain the current transaction data semantic understanding feature vectors.
Accordingly, the current data semantic encoding module 140 includes: and the current data feature extraction unit is used for enabling the current transaction data to pass through a current transaction data feature extractor containing an embedded layer and based on a converter module to obtain the semantic understanding feature vector of the current transaction data. More specifically, the current data feature extraction unit further includes: an embedding transformation subunit, configured to transform each data item in the current transaction data into a plurality of current transaction data item embedding vectors through the embedding layer; and a context encoding subunit, configured to perform context semantic encoding based on a multi-head attention mechanism on the plurality of current transaction data item embedded vectors using the current transaction data feature extractor based on the converter module to obtain the current transaction data semantic understanding feature vector.
In particular, in the technical scheme of the application, the process for generating the wind control result based on the semantic understanding feature vector of the current transaction data and the semantic understanding feature vector of the historical transaction data comprises the following steps: first, a transfer matrix of the semantic understanding feature vector of the current transaction data relative to the semantic understanding feature vector of the historical transaction data is calculated. That is, in the technical solution of the present application, the feature expression of the comparison result between the recent transaction data and the historical transaction pattern of the client in the high-dimensional semantic feature space is represented by the transfer matrix of the semantic understanding feature vector of the current transaction data relative to the semantic understanding feature vector of the historical transaction data. And then, the transfer matrix is passed through a classifier to obtain a classification result as the wind control result, wherein the classification result is used for indicating whether the current transaction is abnormal or not. That is, the classifier is used to determine a class probability tag to which the transition matrix belongs, where the class probability tag is used to indicate whether the current transaction is abnormal.
Accordingly, as shown in fig. 3, the wind control result generating module 150 includes: a transfer matrix calculation unit 151, configured to calculate a transfer matrix of the semantic understanding feature vector of the current transaction data with respect to the semantic understanding feature vector of the historical transaction data; and a classification unit 152, configured to pass the transfer matrix through a classifier to obtain a classification result as the wind control result, where the classification result is used to indicate whether the current transaction is abnormal.
That is, in the technical solution of the present disclosure, the labels of the classifier include that there is an abnormality (first label) of the current transaction, and that there is no abnormality (second label) of the current transaction, wherein the classifier determines to which classification label the transfer matrix belongs through a soft maximum function. It should be noted that the first tag p1 and the second tag p2 do not include a manually set concept, and in fact, during the training process, the computer model does not have a concept of "whether the current transaction is abnormal", which is only two kinds of classification tags, and the probability that the output feature is under the two classification tags, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the current transaction is abnormal is actually converted into a classification probability distribution conforming to the classification of the natural law through classifying the tags, and the physical meaning of the natural probability distribution of the tags is essentially used instead of the language text meaning of whether the current transaction is abnormal.
It should be appreciated that the role of the classifier is to learn the classification rules and classifier using a given class, known training data, and then classify (or predict) the unknown data. Logistic regression (logistics), SVM, etc. are commonly used to solve the classification problem, and for multi-classification problems (multi-class classification), logistic regression or SVM can be used as well, but multiple bi-classifications are required to compose multiple classifications, but this is error-prone and inefficient, and the commonly used multi-classification method is the Softmax classification function.
Accordingly, in one possible implementation manner, the transferring matrix is passed through a classifier to obtain a classification result as the wind control result, where the classification result is used to indicate whether the current transaction is abnormal, and the method includes: expanding the transfer matrix into classification feature vectors according to row vectors or column vectors; performing full-connection coding on the classification feature vectors by using a full-connection layer of the classifier to obtain coded classification feature vectors; and inputting the coding classification feature vector into a Softmax classification function of the classifier to obtain the classification result.
It is worth mentioning that the transfer matrix is a matrix describing the transfer relationship between two sequences. The transfer matrix calculation unit 151 is configured to calculate a transfer matrix of the semantic understanding feature vector of the current transaction data with respect to the semantic understanding feature vector of the historical transaction data. The calculation of the transfer matrix may be implemented by different methods, for example, based on similarity measurement, attention mechanism, etc., which is used to measure the similarity and correlation between the current transaction data and the historical transaction data, so as to capture the evolution and transfer rules of the transaction behavior. By calculating the transfer matrix, the semantic understanding feature vector of the historical transaction data can be compared with the semantic understanding feature vector of the current transaction data, and the transfer relationship between the semantic understanding feature vector and the semantic understanding feature vector of the current transaction data can be analyzed. This helps determine if the current transaction is abnormal because abnormal transactions may have significant differences from the transfer relationship of the historical transaction data. The classification unit 152 uses the transfer matrix as an input, and classifies the transfer matrix by a classifier, resulting in a classification result as an air control result. This classification result indicates whether the current transaction is abnormal, i.e. at risk. The transfer matrix provides information for measuring the transfer relation between the current transaction and the historical transaction, and the risk assessment and wind control decision of the current transaction can be realized by judging the transfer matrix through the classifier. Therefore, the role of the transfer matrix in the wind control result generation module is to measure the transfer relation between the current transaction and the historical transaction, and to carry out risk assessment and wind control decision through the classifier.
Further, the intelligent wind control system for trading clients based on artificial intelligence technology further comprises a training module for training the semantic encoder including an embedded layer, the current transaction data feature extractor based on the converter module including an embedded layer, and the classifier.
More specifically, in one specific example, as shown in fig. 4, the training module 200 includes: a training data obtaining unit 210, configured to obtain training data, where the training data includes training historical transaction data, training current transaction data, and a true value of whether the training current transaction is normal; a historical transaction data training encoding unit 220, configured to perform training encoding on the training historical transaction data by using the semantic encoder including the embedded layer and the current transaction data feature extractor based on the converter module to obtain a training historical transaction data semantic understanding feature vector; a current transaction data training semantic coding unit 230, configured to use the current transaction data feature extractor including the embedded layer and based on the converter module to perform semantic coding on the training current transaction data to obtain a training current transaction data semantic understanding feature vector; a training transfer matrix calculating unit 240, configured to calculate a training transfer matrix of the training current transaction data semantic understanding feature vector relative to the training historical transaction data semantic understanding feature vector; a classification loss function value calculation unit 250, configured to pass the training transfer matrix through a classifier to obtain a classification loss function value; a common manifold implicit similarity factor calculation unit 260, configured to calculate a common manifold implicit similarity factor between the training current transaction data semantic understanding feature vector and the training historical transaction data semantic understanding feature vector; and a loss training unit 270, configured to train the semantic encoder including the embedding layer, the current transaction data feature extractor based on the converter module including the embedding layer, and the classifier with a weighted sum of the classification loss function value and the common manifold implicit similarity factor as a loss function value.
In particular, in the technical scheme of the application, the training historical transaction data semantic understanding feature vector is obtained by performing secondary semantic encoding on the training historical transaction data (namely, performing semantic encoding on each training transaction data first and then performing semantic encoding on semantic encoding feature representation of each training transaction data), and the training current transaction data semantic understanding feature vector is obtained by performing one-time semantic encoding on training current transaction data, so that feature gradient and depth difference exist between the training historical transaction data semantic understanding feature vector and the training current transaction data semantic understanding feature vector. Therefore, in the case of calculating the domain transfer feature of the training current transaction data semantic understanding feature vector relative to the training historical transaction data semantic understanding feature vector, the training transfer matrix also has the problem of poor geometric monotonicity of the high-dimensional feature manifold expression in the high-dimensional feature space, thereby affecting the convergence effect of the classification regression thereof through the classifier, i.e. reducing the training speed and the accuracy of the training result.
Based on this, the applicant of the present application semantically understands feature vectors for the training current transaction data, e.g., noted asAnd the training historical transaction data semantic understanding feature vector, e.g. noted +.>A common manifold implicit similarity factor is introduced as a loss function.
Accordingly, in a specific example, the common manifold implicit similarity factor calculation unit 260 is configured to: calculating the common manifold implicit similarity factor between the training current transaction data semantic understanding feature vector and the training historical transaction data semantic understanding feature vector with a factor calculation formula; wherein, the factor calculation formula is:
wherein ,representing the training current transaction data semantic understanding feature vector,/->Representing the training historical transaction data semantic understanding feature vector, the training current transaction data semantic understanding feature vector +.>And said training history transaction data semantic understanding feature vector +.>Are all in the form of column vectors>Representing a transpose operation->Representing the two norms of the vector, and +.>Representing the square root of the Frobenius norm of the matrix,/i>、/>、/> and />For the weight super parameter, ++>Representing vector multiplication, ++>Representing vector subtraction +.>Representing multiplication by location +.>Representing the common manifold implicit similarity factor.
Here, the common manifold implicit similarity factor may semantically understand feature vectors with the training current transaction dataAnd said training history transaction data semantic understanding feature vector +.>The structural association between the training current transaction data semantic understanding feature vector and the training current transaction data semantic understanding feature vector are shared by the same factorization weight to represent the common manifold of the respective feature manifolds in the cross dimension>And said training history transaction data semantic understanding feature vector +.>And (3) the common constraint of manifold structural factors such as variability, correspondence, relevance and the like, so as to measure the distribution similarity of geometric derivative structure representations depending on common manifold, thereby realizing nonlinear geometric monotonicity of domain transfer characteristics among cross-modal characteristic distributions, improving the geometric monotonicity of high-dimensional characteristic manifold representations of the training transfer matrix, and improving the convergence effect of classification regression of the training transfer matrix through a classifier, namely improving the training speed and the accuracy of training results.
In summary, the intelligent wind control system 100 for trading customers based on the artificial intelligence technology according to the embodiment of the application is illustrated, which can learn the historical trading behavior of customers based on the artificial intelligence technology, and timely identify and follow up the recent abnormal trading behavior.
As described above, the intelligent wind control system 100 for a customer transaction based on the artificial intelligence technology according to the embodiment of the present application may be implemented in various terminal devices, for example, a server having an intelligent wind control algorithm for a customer transaction based on the artificial intelligence technology according to the embodiment of the present application, etc. In one example, the intelligent wind control system 100 for customer transactions based on artificial intelligence techniques in accordance with embodiments of the present application may be integrated into the terminal device as a software module and/or hardware module. For example, the intelligent wind control system 100 for customer transactions based on artificial intelligence techniques according to embodiments of the present application may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the intelligent wind control system 100 based on the artificial intelligence technology for trading customers according to the embodiment of the present application may be one of a plurality of hardware modules of the terminal device.
Alternatively, in another example, the intelligent wind control system 100 for a customer transaction based on the artificial intelligence technology according to the embodiment of the present application may be a separate device from the terminal device, and the intelligent wind control system 100 for a customer transaction based on the artificial intelligence technology may be connected to the terminal device through a wired and/or wireless network and transmit the interactive information according to a agreed data format.
FIG. 5 is a flow chart of an intelligent wind control method for a customer transaction based on artificial intelligence technology in accordance with an embodiment of the present application. FIG. 6 is a schematic diagram of a system architecture of an intelligent wind control method for customer transactions based on artificial intelligence technology according to an embodiment of the present application. As shown in fig. 5 and 6, an intelligent wind control method for a customer transaction based on an artificial intelligence technology according to an embodiment of the present application includes: s110, acquiring historical transaction data of a served object, wherein each transaction data in the historical transaction data comprises transaction time, transaction variety, transaction direction, transaction price, transaction quantity, transaction cost, transaction type, transaction result, transaction state and transactor information; s120, acquiring current transaction data of the served object; s130, carrying out semantic coding on the historical transaction data to obtain a semantic understanding feature vector of the historical transaction data; s140, carrying out semantic coding on the current transaction data to obtain semantic understanding feature vectors of the current transaction data; and S150, generating a wind control result based on the semantic understanding feature vector of the current transaction data and the semantic understanding feature vector of the historical transaction data.
In a specific example, in the intelligent wind control method for client transaction based on artificial intelligence technology, the semantic encoding of the historical transaction data to obtain the semantic understanding feature vector of the historical transaction data includes: each transaction data in the historical transaction data is respectively passed through a semantic encoder comprising an embedded layer to obtain a plurality of transaction data semantic encoding feature vectors; and passing the plurality of transaction data semantically encoded feature vectors through a historical transaction data feature extractor based on a converter module to obtain the historical transaction data semantically understood feature vector.
Here, it will be appreciated by those skilled in the art that the specific operation of the steps in the above-described intelligent wind control method for a customer transaction based on the artificial intelligence technique has been described in detail in the above description of the intelligent wind control system 100 for a customer transaction based on the artificial intelligence technique with reference to fig. 1 to 4, and thus, a repetitive description thereof will be omitted.
FIG. 7 is an application scenario diagram of an intelligent wind control system for customer transactions based on artificial intelligence techniques according to an embodiment of the present application. As shown in fig. 7, in this application scenario, first, historical transaction data (e.g., D1 illustrated in fig. 7) of a served object and current transaction data (e.g., D2 illustrated in fig. 7) of a served object are acquired, wherein each transaction data in the historical transaction data contains transaction time, transaction variety, transaction direction, transaction price, transaction number, transaction fee, transaction type, transaction result, transaction status, and transactor information, and then the historical transaction data and the current transaction data are input into a server (e.g., S illustrated in fig. 7) where an intelligent wind control algorithm for a customer transaction based on an artificial intelligence technique is deployed, wherein the server is capable of processing the historical transaction data and the current transaction data using the intelligent wind control algorithm for a customer transaction based on an artificial intelligence technique to obtain a classification result for indicating whether the current transaction is abnormal.
Further, in another embodiment of the present application, there is also provided an intelligent wind control system for trading and trading systems of clients based on artificial intelligence technology, including: 1. learning the actual routing data order of the historical order, and identifying and early warning the current abnormal overtime order in time; 2. identifying and early warning stock price abnormal fluctuation by learning market trade variety price fluctuation; 3. timely identifying the time delay condition of the market data by learning the market update frequency and time delay, and early warning to an operation and maintenance team; 4. learning the historical transaction behavior of the client, and timely identifying and following the recent abnormal transaction behavior.
The application uses specific words to describe embodiments of the application. Reference to "a first/second embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic is associated with at least one embodiment of the application. Thus, it should be emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various positions in this specification are not necessarily referring to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of the application may be combined as suitable.
Furthermore, those skilled in the art will appreciate that the various aspects of the application are illustrated and described in the context of a number of patentable categories or circumstances, including any novel and useful procedures, machines, products, or materials, or any novel and useful modifications thereof. Accordingly, aspects of the application may be performed entirely by hardware, entirely by software (including firmware, resident software, micro-code, etc.) or by a combination of hardware and software. The above hardware or software may be referred to as a "data block," module, "" engine, "" unit, "" component, "or" system. Furthermore, aspects of the application may take the form of a computer product, comprising computer-readable program code, embodied in one or more computer-readable media.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The foregoing is illustrative of the present application and is not to be construed as limiting thereof. Although a few exemplary embodiments of this application have been described, those skilled in the art will readily appreciate that many modifications are possible in the exemplary embodiments without materially departing from the novel teachings and advantages of this application. Accordingly, all such modifications are intended to be included within the scope of this application as defined in the following claims. It is to be understood that the foregoing is illustrative of the present application and is not to be construed as limited to the specific embodiments disclosed, and that modifications to the disclosed embodiments, as well as other embodiments, are intended to be included within the scope of the appended claims. The application is defined by the claims and their equivalents.
Claims (10)
1. An intelligent wind control system for trading clients based on artificial intelligence technology, comprising:
the historical transaction data acquisition module is used for acquiring historical transaction data of the served object, wherein each transaction data in the historical transaction data comprises transaction time, transaction variety, transaction direction, transaction price, transaction quantity, transaction cost, transaction type, transaction result, transaction state and transactor information;
the current transaction data acquisition module is used for acquiring current transaction data of the served object;
the historical data semantic coding module is used for carrying out semantic coding on the historical transaction data to obtain a historical transaction data semantic understanding feature vector;
the current data semantic coding module is used for carrying out semantic coding on the current transaction data to obtain a semantic understanding feature vector of the current transaction data; and
and the wind control result generation module is used for generating a wind control result based on the semantic understanding feature vector of the current transaction data and the semantic understanding feature vector of the historical transaction data.
2. The intelligent wind control system for trading customers based on artificial intelligence technology of claim 1, wherein the historical data semantic encoding module comprises:
the historical data embedding and encoding unit is used for enabling each transaction data in the historical transaction data to pass through a semantic encoder comprising an embedding layer respectively so as to obtain a plurality of transaction data semantic encoding feature vectors; and
and the historical data feature extraction unit is used for enabling the transaction data semantic coding feature vectors to pass through a historical transaction data feature extractor based on the converter module to obtain the historical transaction data semantic understanding feature vectors.
3. The intelligent wind control system for trading customers based on artificial intelligence technology of claim 2, wherein the current data semantic coding module comprises:
and the current data feature extraction unit is used for enabling the current transaction data to pass through a current transaction data feature extractor containing an embedded layer and based on a converter module to obtain the semantic understanding feature vector of the current transaction data.
4. An intelligent wind control system for trading customers based on artificial intelligence technology as claimed in claim 3, wherein the current data feature extraction unit comprises:
an embedding transformation subunit, configured to transform each data item in the current transaction data into a plurality of current transaction data item embedding vectors through the embedding layer; and
and the context coding subunit is used for carrying out context semantic coding based on a multi-head attention mechanism on the plurality of current transaction data item embedded vectors by using the current transaction data feature extractor based on the converter module so as to obtain the current transaction data semantic understanding feature vector.
5. The intelligent wind control system for trading customers based on artificial intelligence technology of claim 4, wherein the wind control result generation module comprises:
the transfer matrix calculation unit is used for calculating a transfer matrix of the semantic understanding feature vector of the current transaction data relative to the semantic understanding feature vector of the historical transaction data; and
and the classification unit is used for passing the transfer matrix through a classifier to obtain a classification result as the wind control result, wherein the classification result is used for indicating whether the current transaction is abnormal or not.
6. The intelligent wind control system for customer transactions based on artificial intelligence techniques according to claim 5, further comprising a training module for training the semantic encoder including an embedded layer, the current transaction data feature extractor based on a converter module including an embedded layer, and the classifier.
7. The intelligent wind control system for trading customers based on artificial intelligence technology of claim 6, wherein the training module comprises:
the training data acquisition unit is used for acquiring training data, wherein the training data comprises training historical transaction data, training current transaction data and a true value for training whether the current transaction is normal or not;
the historical transaction data training and encoding unit is used for training and encoding the training historical transaction data by using the semantic encoder comprising the embedded layer and the current transaction data feature extractor based on the converter module so as to obtain training historical transaction data semantic understanding feature vectors;
the training semantic coding unit of the current transaction data is used for carrying out semantic coding on the training current transaction data by using the current transaction data feature extractor containing the embedded layer and based on the converter module so as to obtain a training current transaction data semantic understanding feature vector;
the training transfer matrix calculation unit is used for calculating a training transfer matrix of the training current transaction data semantic understanding feature vector relative to the training historical transaction data semantic understanding feature vector;
a classification loss function value calculation unit, configured to pass the training transfer matrix through a classifier to obtain a classification loss function value;
the common manifold implicit similarity factor calculation unit is used for calculating a common manifold implicit similarity factor between the training current transaction data semantic understanding feature vector and the training historical transaction data semantic understanding feature vector; and
and the loss training unit is used for training the semantic encoder containing the embedded layer, the current transaction data characteristic extractor based on the converter module containing the embedded layer and the classifier by taking the weighted sum of the classification loss function value and the common manifold implicit similarity factor as the loss function value.
8. The intelligent wind control system for customer transactions based on artificial intelligence technology according to claim 7, wherein said common manifold implicit similarity factor calculation unit is configured to:
calculating the common manifold implicit similarity factor between the training current transaction data semantic understanding feature vector and the training historical transaction data semantic understanding feature vector with a factor calculation formula;
wherein, the factor calculation formula is:
wherein ,representing the training current transaction data semantic understanding feature vector,/->Representing the training historical transaction data semantic understanding feature vector, the training current transaction data semantic understanding feature vector +.>And said training history transaction data semantic understanding feature vector +.>Are all in the form of column vectors>Representing a transpose operation->Representing the two norms of the vector, anRepresenting the square root of the Frobenius norm of the matrix,/i>、/>、/> and />For the weight super parameter, ++>Representing vector multiplication, ++>Representing vector subtraction +.>Representing multiplication by location +.>Representing the common manifold implicit similarity factor.
9. An intelligent wind control method for trading clients based on artificial intelligence technology is characterized by comprising the following steps:
acquiring historical transaction data of a served object, wherein each transaction data in the historical transaction data comprises transaction time, transaction variety, transaction direction, transaction price, transaction quantity, transaction cost, transaction type, transaction result, transaction state and transactor information;
acquiring current transaction data of a served object;
carrying out semantic coding on the historical transaction data to obtain a semantic understanding feature vector of the historical transaction data;
carrying out semantic coding on the current transaction data to obtain semantic understanding feature vectors of the current transaction data; and
and generating a wind control result based on the current transaction data semantic understanding feature vector and the historical transaction data semantic understanding feature vector.
10. The intelligent wind control method for customer transactions based on artificial intelligence technology according to claim 9, wherein semantically encoding said historical transaction data to obtain a historical transaction data semantic understanding feature vector comprises:
each transaction data in the historical transaction data is respectively passed through a semantic encoder comprising an embedded layer to obtain a plurality of transaction data semantic encoding feature vectors; and
passing the plurality of transaction data semantically encoded feature vectors through a historical transaction data feature extractor based on a converter module to obtain the historical transaction data semantically understood feature vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311101297.6A CN116843345A (en) | 2023-08-30 | 2023-08-30 | Intelligent wind control system and method for trading clients based on artificial intelligence technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311101297.6A CN116843345A (en) | 2023-08-30 | 2023-08-30 | Intelligent wind control system and method for trading clients based on artificial intelligence technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116843345A true CN116843345A (en) | 2023-10-03 |
Family
ID=88162016
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311101297.6A Pending CN116843345A (en) | 2023-08-30 | 2023-08-30 | Intelligent wind control system and method for trading clients based on artificial intelligence technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116843345A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021169364A1 (en) * | 2020-09-23 | 2021-09-02 | 平安科技（深圳）有限公司 | Semantic emotion analysis method and apparatus, device, and storage medium |
| US20210374756A1 (en) * | 2020-05-29 | 2021-12-02 | Mastercard International Incorporated | Methods and systems for generating rules for unseen fraud and credit risks using artificial intelligence |
| US20210374360A1 (en) * | 2020-05-29 | 2021-12-02 | Fmr Llc | Generating Customized Digital Documents Using Artificial Intelligence |
| US20220028376A1 (en) * | 2020-11-18 | 2022-01-27 | Beijing Baidu Netcom Science Technology Co., Ltd. | Method for semantic recognition, electronic device, and storage medium |
| CN115456789A (en) * | 2022-11-10 | 2022-12-09 | 杭州衡泰技术股份有限公司 | Abnormal transaction detection method and system based on transaction pattern recognition |
| CN115983984A (en) * | 2023-02-20 | 2023-04-18 | 杭银消费金融股份有限公司 | Multi-model fusion client risk rating method |
| CN116468250A (en) * | 2023-05-15 | 2023-07-21 | 江西合之信知识产权服务有限公司 | Enterprise research and development project fund supervision system and method thereof |
| CN116658492A (en) * | 2023-07-28 | 2023-08-29 | 新疆塔林投资(集团)有限责任公司 | Intelligent power catwalk and method thereof |
| CN116668380A (en) * | 2023-07-28 | 2023-08-29 | 北京中科网芯科技有限公司 | Message processing method and device of convergence diverter equipment |
-
2023
- 2023-08-30 CN CN202311101297.6A patent/CN116843345A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210374756A1 (en) * | 2020-05-29 | 2021-12-02 | Mastercard International Incorporated | Methods and systems for generating rules for unseen fraud and credit risks using artificial intelligence |
| US20210374360A1 (en) * | 2020-05-29 | 2021-12-02 | Fmr Llc | Generating Customized Digital Documents Using Artificial Intelligence |
| WO2021169364A1 (en) * | 2020-09-23 | 2021-09-02 | 平安科技（深圳）有限公司 | Semantic emotion analysis method and apparatus, device, and storage medium |
| US20220028376A1 (en) * | 2020-11-18 | 2022-01-27 | Beijing Baidu Netcom Science Technology Co., Ltd. | Method for semantic recognition, electronic device, and storage medium |
| CN115456789A (en) * | 2022-11-10 | 2022-12-09 | 杭州衡泰技术股份有限公司 | Abnormal transaction detection method and system based on transaction pattern recognition |
| CN115983984A (en) * | 2023-02-20 | 2023-04-18 | 杭银消费金融股份有限公司 | Multi-model fusion client risk rating method |
| CN116468250A (en) * | 2023-05-15 | 2023-07-21 | 江西合之信知识产权服务有限公司 | Enterprise research and development project fund supervision system and method thereof |
| CN116658492A (en) * | 2023-07-28 | 2023-08-29 | 新疆塔林投资(集团)有限责任公司 | Intelligent power catwalk and method thereof |
| CN116668380A (en) * | 2023-07-28 | 2023-08-29 | 北京中科网芯科技有限公司 | Message processing method and device of convergence diverter equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | A novel deep learning based fault diagnosis approach for chemical process with extended deep belief network | |
| Zhu et al. | Estimation of bearing remaining useful life based on multiscale convolutional neural network | |
| CN109034368B (en) | DNN-based complex equipment multiple fault diagnosis method | |
| Wu et al. | A weighted deep domain adaptation method for industrial fault prognostics according to prior distribution of complex working conditions | |
| Xu et al. | Spatio-temporal degradation modeling and remaining useful life prediction under multiple operating conditions based on attention mechanism and deep learning | |
| CN116340796B (en) | Time sequence data analysis method, device, equipment and storage medium | |
| CN113486578A (en) | Method for predicting residual life of equipment in industrial process | |
| CN116929815A (en) | Equipment working state monitoring system and method based on Internet of things | |
| CN117041017B (en) | Intelligent operation and maintenance management method and system for data center | |
| CN116911929B (en) | Advertisement service terminal and method based on big data | |
| CN111340110A (en) | Fault early warning method based on industrial process running state trend analysis | |
| CN116663568B (en) | Critical task identification system and method based on priority | |
| CN112560997A (en) | Fault recognition model training method, fault recognition method and related device | |
| CN110781970A (en) | Method, device and equipment for generating classifier and storage medium | |
| Kong et al. | A high generalizable feature extraction method using ensemble learning and deep auto-encoders for operational reliability assessment of bearings | |
| Zhang | Application of artificial intelligence recognition technology in digital image processing | |
| Xu et al. | Global attention mechanism based deep learning for remaining useful life prediction of aero-engine | |
| Kim et al. | A convolutional transformer model for multivariate time series prediction | |
| CN117078007A (en) | Multi-scale wind control system integrating scale labels and method thereof | |
| CN117289736A (en) | Temperature control method and system for high-temperature anti-corrosion storage tank | |
| CN117037427B (en) | Geological disaster networking monitoring and early warning system | |
| CN117578715A (en) | Intelligent monitoring and early warning method, system and storage medium for power operation and maintenance | |
| Wang et al. | Predict industrial equipment failure with time windows and transfer learning | |
| CN116843345A (en) | Intelligent wind control system and method for trading clients based on artificial intelligence technology | |
| CN113835964B (en) | Cloud data center server energy consumption prediction method based on small sample learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |